PR(A) = (1-d) + d (PR(T1)/C(T1) + … + PR(Tn)/C(Tn))
· PR(A) – PageRank of the page
· d – Damping Factor, or the probability that a user will keep clicking on links, which is generally assumed to be 0.85
· Tn – Page with a link to A, where n is the page that has the link on it
· C – How many links page Tn has on it
In English, the PageRank of a page is the sum of each page’s PageRank divided by their outbound links, multiplied by d, with 1-d added afterwards. So, if d is still assumed to be 0.85, then a page without anything linking to it should have a PageRank of 0.15 (which is really low, as big websites have at least thousands of points).
In Internet marketing, an entire field is dedicated to making webpages appear at the top of search engine results. This field is SEO, or Search Engine Optimization, and requires extensive knowledge of ranking algorithms such as Google’s PageRank. Because many people use search engines to find information, good optimizers can find themselves in demand by companies that want to get their brand off the ground.
Even though PageRank is just an algorithm, it changed the Internet by changing the way people found information. However, the algorithm is capable of more than just ranking pages. Today, similar weighting methods can be found as a driving force behind road planning, social networking, and data analysis in biology.
Taking a step back, PageRank and classifiers like it can be considered part of link analysis – itself a part of networking theory. Link analysis is a field in which the associations between objects are explored. Employed whenever relationships between many different objects need to be analyzed, link analysis finds use in fraud detection, epidemiology, law enforcement investigations, and, of course, search engines.